the synthesis of anatase tio2 nanoparticles by solvothermal method using ionic liquid as additive

ARTICLE

DOI: 10.1002/zaac.201000073

The Synthesis of Anatase TiO2 Nanoparticles by Solvothermal Method usingIonic Liquid as Additive

Hechun Lin,[a] Peter W. de Oliveira,[a] Ingrid Grobelsek,[a] Aude Haettich,[a] andMichael Veith*[a,b]

Dedicated to Professor Rüdiger Kniep on the Occasion of His 65th Birthday

Keywords: Titanates; Sol-gel processes; Ionic liquids; Anatase; Solvothermal reactions

Abstract. The solvothermal reactions of Ti(OiPr)4 in alcohol usingionic liquid as additive were investigated. In the presence of[BMIM][Cl], [BMIM][Br], [BMIM][NTf2], [BMIM][SO3Me],[BMIM][SO4Me], or [BMIM][OTf] (BMIM = 1-Butyl-3-methylimida-zolium), pure anatase nanoparticles were obtained. The controlled hy-

IntroductionTitanium oxide (TiO2) nanomaterials have received signifi-cant attention because of their numerous applications such asphotocatalysts, pigments, photovoltaic cells, gas sensors, andso on [1]. The physical and chemical properties of TiO2 canbe controlled by its particle size, morphology, and crystallinemodification. Many synthetical methods were used to obtainTiO2 nanomaterials, such as sol-gel, hydrothermal, solvother-mal, sonochemical, chemical vapour deposition, electrodeposi-tion, etc.Room temperature ionic liquids, due to their special proper-ties, such as negligible vapour pressure, low toxicity, low melt-ing points, high chemical and thermal stability and wide elec-trochemical window [2, 3], may be seen as a tool with potentialfor use in sustainable processes such as solvent replacement[4, 5], in catalytic reactions [6, 7], in electrochemistry [8], inbiocatalysis [9–11] and in the synthesis of nanomaterials [12–16]. Much research has been focused on the synthesis of TiO2nanomaterials. Hollow TiO2 amorphous microspheres [17], an-atase nanoparticles [18–28], anatase mesoporous monoliths[29], rutile nanostructures [30–33] and TiO2 (B) in anatase na-nopaticles [34] were synthesised in ionic liquids.The growth of TiO2 nanostructures in ionic liquids is a com-plex process. Ionic liquids are normally considered to act asreaction media, templates or surfactants. Dionysiou et al.

* Prof. Dr. Dr. h. c. M. VeithE-Mail: [email protected]

[a] INM – Leibniz Institute for New MaterialsCampus D2 266123 Saarbrücken, Germany

[b] Institut für Anorganische ChemieUniversität des SaarlandesPostfach 15 11 5066041 Saarbrücken, Germany

Z. Anorg. Allg. Chem. 2010, 636, 1947–1954 © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, WeinheimView this journal online at

wileyonlinelibrary.com 1947

drolysis of Ti(OiPr)4 in the presence of ionic liquids to form titaniumoxo clusters plays a key role in the formation of anatase nanostructures,and ionic liquids can be repeatedly used to synthesise anatase nanopar-ticles. However, in the presence of [BMIM][PF6], [BMIM]2[Ti(OH)6]was obtained by an anion exchange reaction.

pointed out that the use of 1-butyl-3-methyl-imidazolium hexa-flurophosphate ([BMIM][PF6]) in the synthesis of anatase mes-oporous nanoparticles induces controlled hydrolysis of tita-nium alkoxide, but there was no obvious chemical bondformation between [BMIM][PF6] and titanate [19]. However,Liu et al. found the formation of an ionic liquid-stabilised poly-anion complex in the microwave-assisted synthesis of size-controlled anatase nanoparticles using [BMIM][BF4] as me-dium [22]. Yang et al. proposed that the self-organisation abil-ity of [BMIM][BF4] or [BMIM][PF6] leads to anatase crystalli-sation by condensation of the Ti-OH groups of hydroxoxylatedtitanium compound in the thermal annealing process [26]. Yuet al. suggested that Ti4+ ions are all octahedrally coordinatedby ligands like [OH]– and [Cl]– through formation of chloridecomplexes of the type [Ti(OH)mCln]2– (m + n = 6) in the syn-thesis of rutile TiO2 nanorods in [BMIM][Cl] with TiCl4 asprecursor [31]. Using FTIR, Zhai et al. confirmed the forma-tion of a bidentate chelating complexation between the carbox-ylic functional group of 1-methylimidazolium-3-acetate chlo-ride ([AcMIM][Cl]) and titanate in synthesis of rutile TiO2[32]. Most recently, Zheng et al. proposed that the interactionbetween imidazolium cation and TiO6 octahedra could be adecisive factor for the formation of the rutile phase, which isbeneficial for the catenarian nuclei by edge-sharing polycon-densation between TiO6 octahedra due to the mutual π-stakingbetween aromatic rings [33]. HF and H2SO4 were used as addi-tives to inhibit the phase transformation in hydrothermal syn-thesis of anatase nanocrystals assisted by ionic liquid [25].In a recent article, we reported about the early hydrolysisstages of Ti(OiPr)4 in imidazolium based ionic liquids/ethanolmixtures [35]. We found that most ionic liquids such as[BMIM][Cl], [BMIM][Br], [BMIM][PF6], [BMIM][N(Tf)2](Tf: SO3CF3), [BMIM][OTf] (Tf: COCF3) and [BMIM]-

H. Lin, P. W. de Oliveira, I. Grobelsek, A. Haettich, M. VeithARTICLE

Scheme 1.

[NCN2], etc, as additives controlled the hydrolysis of Ti(OiPr)4to form Ti7O4(OEt)20. There was no obvious chemical bondingbetween ionic liquid and titanium complex as indicated byNMR spectroscopy. In contrast, an anion exchange reaction in[BF4]– based ionic liquids to form hexahydroxotitanate saltswas found (Scheme 1). In this paper, we describe the furtherpolycondensation process under solvothermal conditions [36–47]. The effects of water concentration, different alcohols, andionic liquids on the products were investigated in detail. Pureanatase nanoparticles were obtained, and some ionic liquidscan be used again up to five times to give rise to anatase inquantitative yields.

Experimental SectionAll chemicals were used as received. Ti(OiPr)4 was distilled undervacuum before use. [BMIM][Cl], [BMIM][Br], [BMIM][PF6],[BMIM][BF4], [BMIM][NTf2] were synthesised according to literaturemethod [48]. [BMIM][SO3Me], [BMIM][SO4Me], and [BMIM][OTf]were received from IoLiTec GmbH. Water contents of alcohol weretitrated by Karl Fischer method.

Ti(OiPr)4 (2.84 g, 10 mmol) was added dropwise to the solution ofionic liquid (20 mmol) in ethanol (26 g) at room temperature. Stirringwas performed at room temperature and it was heated under reflux for4 hours to form Ti7O4(OEt)20. Subsequently, the solution was trans-ferred to a 250 mL Teflon flask and cured in an autoclave or, withoutpre-curing process, the mixture was cured directly in the autoclave.The temperature was raised up at a ramp rate of 4.4 °C·min–1 to210 °C. The whole hydrothermal process took 2.75 hours or 12 hours.The solution was cooled down gradually to room temperature in theautoclave. The precipitated solids were separated by centrifugation.The ionic liquid part can be re-used by rotationally evaporating off thelow boiling component under reduced pressure (10–2 atm). The pro-ducts were washed five times with ethanol and were freeze-dried under10–2 atm using water as solvent. There were no obvious absorptionbands of ionic liquids in the FTIR spectra in most of the samples. Thereactions in other alcohols such as propanol, butanol, or iso-propanolwere run similarly to this procedure. For comparison, Ti(OiPr)4 wasalso cured in the absence of ionic liquids. The samples were labelledas SN where N is a progressive integer.

Ti7O4(OEt)20 was synthesised using [BMIM][Cl] as additive. Ti(OiPr)4(2.84 g, 10 mmol) was added dropwise to the solution of [BMIM][Cl](3.49 g, 20 mmol) in ethanol (11.4 g, less 0.1 % water) under magneticstirring at room temperature in an open flask. After stirring for 1 hourat room temperature to obtain a turbid solution, the temperature wasraised up to reflux to solve all precipitates and to obtain a clear liquid.Stirring was continued for 4 hours and afterwards the low boiling com-ponents were distilled off. The liquid was redissolved in CH3CN andEtOAc under reflux and filtered with a fine filter paper to give rise tothe mother liquor. This was kept at room temperature overnight. Col-ourless crystals precipitated, which were grown further in a refrigeratorat –30 °C for one day. Finally, the mother liquor was decanted off. The

1948 www.zaac.wiley-vch.de © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Z. Anorg. Allg. Chem. 2010, 1947–1954

solids were washed with CH3CN for three times and dried with N2 flowto give rise to 1.27 g (70 %) Ti7O4(OEt)20. Subsequently, the obtainedTi7O4(OEt)20 was solved in ethanol (26 g, with less than 0.1 % H2O or2 % H2O, respectively) and was cured in an autoclave for 12 hours at210 °C to give rise to anatase nanoparticles.

Powder X-ray diffraction (XRD) patterns were collected with a BrukerAXS D8 powder diffractometer unit, using Cu-Kα radiation (λ = 0.154nm) operating at 40 kV and 40 mA. A position sensitive detector(LynxEye) based on Bruker AXS' compound silicon strip technologywas applied. The patterns were recorded from 3° to 90° in 2θ with a2θ scan step size of 0.02°. For clarity, the low angle region of pureanatase samples is not shown in the figures. For all powders glass sam-ple holders were used. The structural refinement of anatase and profileanalysis of related powder diffraction patterns were carried out with theprogram TOPAS [49]. In this program, the line profile shapes are de-scribed by convoluting the X-ray source emission profile with the in-strument aberrations and physical properties of the sample (fundamentalparameters convolution-based approach) [50]. The mean crystallitesizes were calculated by the Scherrer equation (Scherrer constant k =1). The integral broadness-based volume-weighted calculation assum-ing intermediate crystallite-size broadening was modelled by a Voigtfunction. Transmission electron microscope (TEM) and high-resolutiontransmission electron microscope pictures (HRTEM) were obtained byusing a JEM-3010 electron microscope. Nitrogen porosimetry was per-formed with a Quantuchrome Autosorb6B Instrument. Surface areaswere calculated using the Brunauer–Emmett–Teller (BET) equation.FTIR spectra were measured with a Bruker Tensor 27 Fourier Trans-form Infrared spectrometer. Raman spectra were obtained with a LabRam Aramis (Horiba Jobin Yvon) spectrometer.

Results and DiscussionSynthesis of TiO2 in [BMIM][Cl] using EtOH as Solvent

Ti(OiPr)4 was firstly hydrolysed in the presence of 2 equiv.[BMIM][Cl] with ethanol as solvent in the presence of lessthan 0.1 % water in the alcohol. After stirring at room temper-ature for 1 hour and heating under reflux for 4 hours, a clearsolution was obtained and Ti7O4(OEt)20 was formed. The mix-ture was transferred to a Teflon flask and hydrothermally curedfor 2.75 hours to give rise to white solids of sample S1. Be-sides two peaks at 2θ values of 8.25° and 16.21°, the formedproduct showed a powder X-ray diffraction (PXRD) patternsimilar to anatase of low range order (Figure 1b-S1) with meancrystallite sizes between 2 and 4 nm. The two reflections atlow angles could not be assigned to any available JCPDS(Joint Committee on Powder Diffraction Standards) pattern.They were also different to the result of Smarsly et al., whoobtained amorphous TiO2 with an additional signal at about2θ = 12o during the synthesis of rutile TiO2 by extraction meth-ods with iPrOH [30]. In the TEM image no crystalline frac-tions could be identified (Figure 2a). The sample was also ana-

Synthesis of Anatase TiO2 Nanoparticles using Ionic Liquid as Additive

lysed with FTIR and Raman spectroscopy. In the FTIRspectrum (Figure 3a), the absorption bands at 1111, 1165,1382, 1465, 1569, 2873, 2963, 3107, 3147 cm–1 belong to

Figure 1. Powder X-ray diffraction (PXRD) patterns of TiO2 particlessynthesised at various conditions. S1: Ti(OiPr)4 (0.01 mol),[BMIM][Cl] (0.02 mol), ethanol (26 g), heated under reflux for 4 h,afterwards cured for 2.75 h at 210 °C in the autoclave. S2: Ti(OiPr)4(0.01 mol), [BMIM][Cl] (0.02 mol), ethanol (26 g), heated under re-flux for 4 h, afterwards cured for 12 h at 210 °C in the autoclave. S3:Ti(OiPr)4 (0.01 mol), [BMIM][Cl] (0.02 mol), ethanol (26 g), curedfor 12 h at 210 °C in the autoclave. S4: Ti(OiPr)4 (0.01 mol),[BMIM][Cl] (0.005 mol), ethanol (26 g), cured for 12 h at 210 °C inthe autoclave. S5: Ti(OiPr)4 (0.01 mol), [BMIM][Cl] (0.01 mol), etha-nol (26 g), cured for 12 h at 210 °C in the autoclave. S6: Ti(OiPr)4(0.01 mol), [BMIM][Cl] (0.02 mol), ethanol (26 g, 2 % water), curedfor 12 h at 210 °C in the autoclave. S7: Ti(OiPr)4 (0.01 mol),[BMIM][Cl] (0.02 mol), ethanol (26 g, 5 % water), cured for 12 h at210 °C in the autoclave.

Figure 2. TEM images of TiO2 particles in different samples S1–S5. (a) S1, (b) S2, (c) S3, (d) S4, (e) S5, (f) HRTEM of S3. Insets are selectedarea electron diffraction patterns.

Z. Anorg. Allg. Chem. 2010, 1947–1954 © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.zaac.wiley-vch.de 1949

[BMIM]+. However, the absorption at 3107 and 3147 cm–1

were different from that in bulk [BMIM][Cl] [51]. The bandcentred around 3257 cm–1 is assigned to the hydroxyl groupof Ti–O–H, and the absorption centred around 1642 cm–1 isassociated with the deformation vibration of H–O–H bondsfrom the physico-sorbed water [52]. In the Raman spectrum(Figure 3b), there is no obvious typical absorption of anatase,however, the absorption bands at 163, 185, 272, 284, 384, 447,673, 700, 825 cm–1 are similar to those of titanate [53]. Theseresults indicate the formation of the titanate and the interactionbetween the imidazolium ring and the titanate. By prolongingthe curing time to 12 hours, crystallised TiO2 (anatase) wasobtained with a mean crystallite size of 9 nm. The diffractionpattern shows reflections of anatase [54] (Figure 1a, S2). Nopeaks for any other phases were detected in the pattern. TEMimage of sample S2 (Figure 2b) shows that almost sphericaland accordingly cube-like particles were formed in agreementwith the ratio of the calculated column heights (crystallite sizebroadening) of the (004):(020) X-ray reflections of 7:8. If themixture was cured directly under the solvothermal conditionswithout the pre-curing process, the resulted powder was simi-lar to that of sample S2 but with a slightly lower mean crystal-lite size of 6 nm (Figure 1a-S3, Figure 2c). The presence of2 equiv. [BMIM][Cl] was a prerequisite for the formation ofcrystalline monophasic anatase nanoparticles. In the presenceof 0.5 equiv. (sample S4) or 1 equiv. [BMIM][Cl] (sample S5),similarly to sample S1 only porous products with broad andweak X-ray reflections with non-identifiable peaks at 2θ valuesof ~7.97° and ~15.77° were obtained (Figure 1b-S4, 5, TEMimage: Figure 2d, e).In the next steps, we checked the effect of water concentra-tions in alcohols to the synthesis. Increasing the water content


Figure 3. (a) FTIR spectrum of sample S1; (b) Raman spectrum of sample S1.

to 2 % (S6) or 5 % (S7) produced single-phase nanoscaledanatase powders with a slightly bigger particle size (see Ta-ble 1) than in case of the samples S2 and S3. As can be seenfrom the TEM images in Figure 4, the presence of 2 % watergave rise to cube-like anatase nanoparticles with a calculatedcrystallite size-ratio of (004):(020) planes of nearly 8:8 inSample S6 and 9:10 in Sample S7. The particles shown inthe HRTEM image are placed on copper grid parallel to [101]direction. The surface area (SBET) was estimated by BET

Table 1. Preparation conditions, crystallite sizes, and BET surface ar-eas of TiO2.a).

Sample water contents in alcohol crystallite size /nm SBET /m2·g–1

S1 < 0.1 %b) – –S2 < 0.1 %c) 9 143S3 < 0.1 %d) 6 163S4 < 0.1 %d),e) – –S5 < 0.1 %d),f) – –S6 2 %d) 10 123S7 5 %d) 11 120

a) Ti(OiPr)4 (0.01 mol), [BMIM][Cl] (0.02 mol), ethanol (26 g) b)Heated under refluxed for 4 h, afterwards cured for 2.75 h at 210 °Cin the autoclave. c) Heated under refluxed for 4 h, afterwards curedfor 12 h at 210 °C in the autoclave. d) Without pre-curing process,cured directly for 12 h at 210 °C in the autoclave e) [BMIM][Cl]:0.005 mol. f) [BMIM][Cl]: 0.01 mol.

Figure 4. TEM images of different TiO2 samples S6 and S7. (a) S6; (b) HRTEM of S6; (c) S7. Inset is a selected area electron diffraction pattern.


method. The SBET is found in a range from 120 to 163 m2·g–1(Table 1). In agreement with the X-ray data and TEM images,the increased concentration of H2O led to a lower surface area.Crystallite sizes and BET surface areas of the Samples S1–S7in dependency of preparation conditions are listed in Table 1.

Synthesis of Anatase in [BMIM][Cl] using other Alcoholsas Solvents

Many different alkoxyl titanium clusters were reported in theliterature [55, 56]. It seemed interesting to see if the additionof ionic liquids has an impact on the hydrolysis of Ti(OiPr)4 inother alcohols to form different titanium clusters. We thereforehydrolysed Ti(OiPr)4 in alcohols such as PrOH, iPrOH, orBuOH. Titanium oxo clusters were successfully formed as de-rived from IR spectroscopy, but an exact specification of thestructure was not possible since no crystals suitable for thesingle-crystal X-ray diffraction analysis were obtained. Wealso performed the solvothermal reactions without the pre-cur-ing step. Using PrOH or iPrOH as solvent with less than 0.1 %water produced, besides anatase at least one more phase withbroad X-ray reflections amongst others at 2θ values of 8.01°and 15.95°. As can be seen from the X-ray patterns of Fig-ure 5b, if Pr(OH) was used, the main phase in the end productwas anatase (S8). In the case of iPr(OH) (S9) a high content ofa non-identified compound was obtained. Increasing the water


content to 2 %, monophasic anatase nanoparticles wereformed; the PXRD patterns are summarised in Figure 5a.PrOH or BuOH (samples S10 and S12) gave rise to mean crys-tallite sizes of 9 nm and 8 nm of anatase, respectively. Thedifference of the column heights ratio of the (004):(020) reflec-tions of 6:8 in S10 and 5:7 in S12 leads to almost sphericalparticles. However, in the presence of iPrOH, bipyramidal na-noparticles were obtained (Figure 6b, d) with a mean crystal-lite size of 13 nm and a crystallite size ratio of (004):(020)reflexes of 20:8. As shown in the HTEM image in Figure 6d,

Figure 5. Powder X-ray diffraction (PXRD) patterns of TiO2 particlessynthesised at various conditions (samples S8–S12). Ti(OiPr)4 (0.01mol), [BMIM][Cl] (0.02 mol), alcohol (26 g), cured for 12 h at 210 °Cin the autoclave. S8: PrOH (26 g); S9: iPrOH (26 g); S10: PrOH (26 g,2 % water); S11: iPrOH(26 g, 2 % water); S12: BuOH(26 g, 2 % water).

Figure 6. TEM images of TiO2 particles from samples S10-S12. (a) S10;(b) S11; (c) S12; (d) HTEM image of S11. Insets are selected area elec-tron diffraction patterns.


the particles are oriented on the copper grid parallel to [101]direction. The results of the different preparation characteris-tics are summarised in Table 2.

Table 2. Preparation conditions, cystallite sizes, and BET surface areasof TiO2.a).

Sample alcohols water contents Crystallite size /nm SBET /m2·g–1

S8 PrOH < 0.1 % 8 –S9 iPrOH < 0.1 % – –S10 PrOH 2 % 9 134S11 iPrOH 2 % 13 102S12 BuOH 2 % 8 146

a) Ti(OiPr)4 (0.01 mol), [BMIM][Cl] (0.02 mol), alcohol (26 g), with-out the pre-curing process, the mixtures were cured directly for 12 hat 210 °C in the autoclave.

Synthesis of TiO2 in other Ionic Liquids

Finally, we investigated the hydrothermal reactions ofTi(OiPr)4 with other ionic liquids. As can be seen from theabove results, water concentration strongly influences the endproduct. 2 % of water content is enough for the complete poly-condensation of the titanium precursors. Herein, we used EtOHcontaining 2 % water as solvent. Without the pre-curing proce-dure, the mixture was cured directly under the solvothermalcondition. The results are shown in Table 3 and the corre-spondent PXRD patterns and TEM images are assembled inFigure 7 and Figure 8, respectively. In the presence of[BMIM][Br], [BMIM][NTf2], [BMIM][SO3Me], [BMIM]-[SO4Me], or [BMIM][OTf], which mediated the hydrolysis ofTi(OiPr)4 to form intermediate Ti7O4(OEt)20 under reflux con-ditions, pure crystalline single-phase anatase nanoparticleswere generated. The different crystallite sizes achieved arelisted in Table 3. In the presence of [BMIM][BF4] (S19),which was found to form [BMIM]2[Ti(OH)6] through anionexchange reaction under reflux conditions, only a smallamount of anatase was obtained with a mean crystallite size of9 nm. The smallest sizes, however with a mean diameter of4 nm, had anatase particles generated in presence of[BMIM][SO4Me]. Surprisingly, in the presence of[BMIM][PF6], although Ti7O4(OEt)20 was obtained under re-flux conditions, no solid precipitated but a biphasic liquid wasobtained. The low boiling point components were evaporatedto give rise to a liquid, which was analysed with NMR spectro-scopy to show a new imidazolium compound. The product wasseparated by re-crystallisation from CH3CN/EtOAC for twotimes. The obtained needle-like crystals were identified as[BMIM]2[Ti(OH)6] using single-crystal X-ray diffraction [35].For comparison, the Ti(OiPr)4 was firstly hydrolysed under re-flux conditions to form Ti7O4(OEt)20, and afterwards cured un-der solvothermal conditions. Ti7O4(OEt)20 was further hydro-lysed to carry out the anion exchange reaction to form[BMIM]2[[Ti(OH)6] as well. These results indicated that[BMIM]2[[Ti(OH)6] was thermally more stable than[BMIM][PF6]. The reason for this anion exchange reaction un-der solvothermal conditions might have to do with the lessstable [PF6]– anion, which could loose fluoride [57].


Table 3. Preparation conditions, crystalline sizes, and BET surface ar-eas of TiO2.a).

Sample Ionic liquids Crystallite size /nm SBET area /m2·g–1

S13 [BMIM][Br] 11 123S14 [BMIM][NTf2] 10 119S15 [BMIM][SO3Me] 8 148S16 [BMIM][SO4Me] 4 179S17 [BMIM][OTf] 10 128S18 [BMIM][PF6] – –S19 [BMIM][BF4] 7 –

a) Ti(OiPr)4 (0.01 mol), ionic liquid (0.02 mol), ethanol (26 g, 2 %H2O). The mixtures were cured directly for 12 h at 210 °C in theautoclave.

As can be seen from the above results, most of the testedionic liquids gave rise to pure anatase nanoparticles, except[BMIM][PF6] and [BMIM][BF4] which both have complexfluoride anions and thus are susceptible to loose fluoride in thereaction. The F– ion was found in these two reaction mixturesby F19 NMR analysis. For comparison, we carried out the samesolvothermal reaction of Ti(OiPr)4 in the absence of ionic liq-uids. In this case, white powders were obtained. However, theyhad a lower crystallinity compared to the products in the pres-ence of ionic liquids as shown in the XRD pattern (Figure 9,S20) and TEM image (Figure 10a). We have found thatTi7O4(OEt)20 can be obtained in good yields by the controlledhydrolysis of Ti(OiPr)4 using ionic liquids as further compo-nents. To know if titanium oxo clusters play a key role in theformation of nanostructures, Ti7O4(OEt)20 was separated bycrystallisation and introduced to the solvothermal reaction af-terwards. Using EtOH with less than 0.1 % H2O (S21) or 2 %H2O (S22) as solvent gave rise to pure crystalline single-phaseanatase nanoparticles as well. The correspondent PXRD-pat-terns and TEM images are shown in Figure 9 and Figure 10,

Figure 7. Powder X-ray diffraction (PXRD) patterns of TiO2 particles synthesised at various conditions. Ti(OiPr)4 (0.01 mol), ionic liquids (0.02mol), EtOH (26 g, 2 % H2O), cured for 12 h at 210 °C in the autoclave. S13: [BMIM][Br]; S14: [BMIM][NTf2]; S15: [BMIM][SO3Me]; S16:[BMIM][SO4Me]; S17: [BMIM][OTf]; S19: [BMIM][BF4].


respectively. With a lower water content in the alcohol, theresulting crystallite size of the particles was nearly half (6 nm)of that at higher water content (11 nm). The physical charac-teristics of the so obtained TiO2 powders are summarised inTable 4. They had similar physical characteristics to S3 andS6 (see before). Taking into account all our results it seemsclear that the controlled hydrolysis of Ti(OiPr)4 by ionic liq-uids passes through the formation of intermediateTi7O4(OEt)20. Thus, this compound plays a key role in the for-mation of anatase nanostructures. The more condensed tita-nium clusters seem to improve the formation of the anatasenanostructure. Water concentration also has effect on the finalproducts. The system itself cannot produce water. It needs suf-ficient water to let the precursor polycondense. When it hasless amount of water, the reaction gives rise to smaller crystal-lite size TiO2 or only leads to partially condensed products.2 % water is enough for polycondensation observed in our set-up.

Re-use of Ionic Liquids

Many TiO2 nanomaterials synthesis methods using ionic liq-uids as media were reported. However, there were only rarereports about the repeated use of ionic liquids in the synthesisof TiO2 until now. This may be due to the fact that it is difficultto separate the ionic liquids from the reaction system or ionicliquids may have been decomposed during the reactions. Inour method, ionic liquids such as [BMIM][Cl], [BMIM][Br],or [BMIM][NTf2] were stable under solvothermal conditions.After the end of the process, the mixture includes TiO2 (ana-tase), ionic liquids, and alcohols. TiO2 can be easily separatedby centrifugation and ionic liquids can be recollected for fur-ther use. For example we have found that [BMIM][Cl] and[BMIM][NTf2] can be repeatedly used up to five times with


Figure 8. TEM images of TiO2 particles. (a) S13; (b) S14; (c) S15; (d) S16; (e) S17; (f) S18. Insets are selected area electron diffraction patterns.

Figure 9. Powder X-ray diffraction (PXRD) patterns of TiO2 particlessynthesised at various conditions. The mixtures were cured directly for12 h at 210 °C in the autoclave. S20: Ti(OiPr)4 (2.84 g), EtOH (26 g);S21: Ti7O4(OEt)20 (1.27 g), EtOH (26 g). S22: Ti7O4(OEt)20 (1.27 g),EtOH (26 g, 2 % H2O).

ethanol (2 % water) as solvent to give rise to pure anatasenanoparticles in quantitative yields.

ConclusionsWe carried out solvothermal reactions of Ti(OiPr)4 in alcoholusing ionic liquid as additive. In the presence of [BMIM][Cl],we performed solvothermal reactions in different water con-centration and different alcohols. The system itself cannot pro-duce water. 2 % water was enough for the complete polycon-densation of the precursor. Spheric or cube-like anatase


Figure 10. TEM images of TiO2 particles. (a) S20; (b) S21; (c) S22;(d) HRTEM of S22. Insets are selected area electron diffraction pat-terns.

nanoparticles were obtained; except the reaction in PrOH gaverise to bipyramidal nanoparticles. We also investigated the ef-fect of different cations to the reactions. Using EtOH (2 %water) as solvent, most ionic liquids gave rise to anatase nano-particles with a small variation of crystallite size. Two excep-tions were the reactions in [BMIM][BF4] or [BMIM][PF6]. Theformer carried out the anion exchange reaction under refluxconditions to form [BMIM]2[Ti(OH)6], and the latter led to the


Table 4. Preparation conditions, crystallite sizes, and BET surface ar-eas of TiO2.a).

Sample precusor Water contents Particle size SBET area /m2·g–1

S20 Ti(OiPr)4 <0.1 % 4 –S21 Ti7O4(OEt)20 <0.1 % 6 144S22 Ti7O4(OEt)20 2 % 11 153

a) The mixtures were cured directly for 12 h at 210 °C in the autoclave.S20: Ti(OiPr)4 (0.01 mol), ethanol (26 g); S21: Ti7O4(OEt)20 (1.27 g),ethanol (26 g); S22: Ti7O4(OEt)20 (1.27 g), ethanol (26 g, 2 % H2O).

same reaction under solvothermal conditions. The controlledhydrolysis of Ti(OiPr)4 in the presence of ionic liquids to formtitanium oxo clusters plays a key role in the formation of ana-tase nanostructures. The more condensed titanium oxo clustersmay improve the formation of the nanostructures. Comparedto other ionic liquids mediating procedures, our methods haveadvantages of using fewer amounts of ionic liquids (2 equiv.)and ionic liquids can be repeatedly used. It is worth noting thationic liquids are normally considered as stable media or asinnocent solvent. Our results indicate that not all ionic liquidswere stable in the reaction system. Some may lead to reactionsplaying a prominent role for the formation of TiO2 nanostruc-tures. The application of TiO2 nanoparticles in optical coatingand the solvothermal reactions of Zr(OPr)4 in ionic liquids tosynthesise ZrO2 nanoparticles are going on in our lab.

AcknowledgementThe authors thank the State of Saarland and the Fonds der ChemischenIndustrie for financial support, Dr. Petra Herbeck-Engel for the meas-urement of FTIR and Raman spectrum, and Dr. Karsten Moh and Dr.Herbert Schmid for the measurement of HTEM.

References[1] X. Chen, S. S. Mao, Chem. Rev. 2007, 107, 2891.[2] Ionic Liquids in Synthesis (Eds.: P. Wasserscheid, T. Welton),

Wiley VCH, Weinheim, Germany 2003.[3] P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112, 3926;

Angew. Chem. Int. Ed. 2000, 39, 3772.[4] M. Avalos, R. Babiano, P. Cintas, J. L. Jimenez, J. C. Palacios,

Angew. Chem. 2006, 118, 4008; Angew. Chem. Int. Ed. 2006, 45,3904.

[5] J. Ranke, S. Stolte, R. Stormann, J. Aming, B. Jastorff, Chem.Rev. 2007, 107, 2183.

[6] V. I. Parvulescu, C. Hardacre, Chem. Rev. 2007, 107, 2615.[7] M. A. P. Martins, C. P. Frizzo, D. N. Moreiro, N. Zanatta, H. G.

Bonacorso, Chem. Rev. 2008, 108, 2015.[8] P. Hapiot, C. Lagrost, Chem. Rev. 2008, 108, 2238.[9] F. von Rantwijk, R. A. Sheldon, Chem. Rev. 2007, 107, 277.[10] P. Dominguez De Maria, Angew. Chem. 2008, 120, 7066; Angew.

Chem. Int. Ed. 2008, 47, 5434.[11] J. C. Plaquevent, J. Levillain, F. Guillen, C. Malhiac, A. C. Gau-

mont, Chem. Rev. 2008, 108, 5035.[12] M. Antonietti, D. Kuang, B. Smarsly, Y. Zhou, Angew. Chem.

2004, 116, 5096; Angew. Chem. Int. Ed. 2004, 43, 4988.[13] R. E. Morris, Angew. Chem. 2008, 120, 450; Angew. Chem. Int.

Ed. 2008, 47, 442.[14] A. Taubert, Z. Li, Dalton Trans. 2007, 723.[15] Z. Li, Z. Jia, Y. Luan, T. Mu, Curr. Op. Solid Stat. Mater. Sci.

2008, 12, 1.


[16] T. L. Greaves, C. J. Drummond, Chem. Soc. Rev. 2008, 37, 1709.[17] T. Nakashima, N. Kimizuka, J. Am. Chem. Soc. 2003, 125, 6386.[18] K. Yoo, H. Choi, D. D. Dionysiou, Chem. Commun. 2004, 2000.[19] H. Choi, Y. Kim, R. S. Varma, D. D. Dionysiou, Chem. Mater.

2006, 18, 5377.[20] E. H. Choi, S. Hong, D. J. Moon, Cata. Lett. 2008, 123, 84.[21] Y. Zhou, M. Antonietti, J. Am. Chem. Soc. 2003, 125, 14960.[22] K. Ding, Z. Miao, Z. Liu, Z. Zhang, B. Han, G. An, S. Miao, Y.

Xie, J. Am. Chem. Soc. 2007, 129, 6362.[23] H. Liu, Y. Liang, H. Hu, M. Wang, Solid Stat. Sci. 2009, 11, 1655.[24] T. Alammar, A. Birkner, O. Shekhah, A. Mudring, Mater. Chem.

Phys. 2010, 120, 109.[25] K. Ding, Z. Miao, B. Hu, G. An, Z. Sun, B. Han, Z. Liu, Lang-

muir 2010, 26, 5129.[26] Y. Liu, M. Chang, H. Shao, M. Huang, A. Yang, J. Mater. Sci.

2010, 45, 369.[27] S. Hu, A. Wang, X. Li, Y. Wang, H. Loewe, Chem. Asian J. 2010,

5, 1171.[28] H. Liu, Y. Liang, H. Hu, M. Wang, Solid Stat. Sci. 2009, 11, 1655.[29] Y. Liu, J. Li, M. Wang, Z. Li, H. Liu, P. He, X. Yang, J. Li, Cry.

Grow. Ses. 2005, 5, 1643.[30] H. Kaper, F. Endres, I. Djerdj, M. Antonietti, B. M. Smarsly, J.

Maier, Y. S. Hu, Small 2007, 3, 1753.[31] N. Yu, L. Gong, H. Song, Y. Liu, D. Yin, J. Solid Stat. Chem.

2007, 180, 799.[32] Y. Zhai, Q. Zhang, F. Liu, G. Gao, Mater. Lett. 2008, 62, 4563.[33] W. Zheng, X. Liu, Z. Yan, L. Zhu, ACS Nano 2009, 3, 115.[34] H. Kaper, Sallard, S.; I. Djerdj, M. Antonietti, B. M. Smarsly.

Chem. Mater. 2010, DOI:10.1021/cm100627g.[35] H. Lin, P. W. de Oliveira, V. Huch, M. Veith, in preparation.[36] X. Li, Q. Peng, J. Yi, X. Wang, Y. Li, Chem. Eur. J. 2006, 12,

2383.[37] B. Wen, C. Liu, Y. Liu, Inorg. Chem. 2005, 44, 6503.[38] B. Wen, C. Liu, Y. Liu, New J. Chem. 2005, 29, 969.[39] B. Wen, C. Liu, Y. Liu, J. Phys. Chem. B 2005, 109, 12372.[40] C. Kim, B. K. Moon, J. Park, B. Choi, H. Seo, J. Cryst. Growth

2003, 257, 309.[41] C. Kim, B. K. Moon, J. Park, S. T. Chung, S. Son, J. Cryst.

Growth 2003, 254, 405.[42] S. Yang, L. Gao, Mater. Chem. Phys. 2006, 99, 437.[43] Y. Wu, H. Liu, B. Xu, Appl. Organomet. Chem. 2007, 21, 146.[44] J. Liu, M. Dong, S. Zuo, Y. Yu, Appl. Clay Sci. 2009, 43, 156.[45] H. C. Yang, G. Liu, Z. Q. Shi, H. S. Chen, Y. G. Jin, S. C. Smith,

J. Zou, H. M. Chen, G. Q. Liu, J. Am. Chem. Soc. 2009, 131,4079.

[46] R. K. Wahi, Y. Liu, J. C. Falkner, V. L. Colvin, J. Coll. Inter. Sci.2006, 302, 530.

[47] J. Liao, L. Shi, S. Yuan, Y. Zhao, J. Fang, J. Phys. Chem. C 2009,113, 18778.

[48] X. Creary, E. D. Wilis, Org. Synth. 2005, 82, 166.[49] General profile and structure analysis software for powder diffrac-

tion data, Bruker Analytical X-ray Systems GmbH, 76187 Karl-sruhe, Germany.

[50] R. W. Cheary, A. Coelho, J. Appl. Crystallogr. 1992, 25, 109.[51] Z. Zhou, T. Wang, H. Xing, Ind. Eng. Chem. Res. 2006, 45, 525.[52] Y. V. Kolen'ko, K. A. Kovnir, A. I. Gavrilov, A. V. Garshev, J.

Frantti, O. I. Lebedev, B. R. Churagulov, G. V. Tendeloo, M. Yo-shimura, J. Phys. Chem. B 2006, 110, 4030.

[53] G. Li, L. Li, J. Boerio-Goates, B. F. Woodfield, J. Am. Chem.Soc. 2005, 127, 8659.

[54] Joint Committee on Powder Diffraction Standards (JCPDS, CardNo. 01–084–1285).

[55] U. Schubert, J. Mater. Chem. 2005, 15, 3701.[56] L. Rozes, N. Steunon, G. Fornasieri, C. Sanchez, Monatsh. Chem.

2006, 137, 501.[57] F. Endres, Phys. Chem. Chem. Phys. 2010, 12, 1648.

Received: February 3, 2010Published Online: July 16, 2010

the synthesis of anatase tio2 nanoparticles by solvothermal method using ionic liquid as additive

Documents